Title: |  | Optimization of Object Extraction Based on One User-Prepared Sample |
| | |
Presenter: | | Shahryar Rahnamayan |
Presenter's Affiliation: | | Ph.D. Candidate, Faculty of Engineering, Pattern Analysis and Machine Intelligence Lab, University of Waterloo, CANADA |
Presenter's E-mail address: | | s2rahnam@engmail.uwaterloo.ca |
Authors: | | S. Rahnamayan, H.R. Tizhoosh, M.M.A. Salama |
Abstract (100 words or less): | | Knowledge- and sample-based learning approaches play a pivotal role in image processing. However, the acquisition and integration of expert knowledge (for the former) and providing a sufficiently large number of training samples (for the latter) are generally hard to perform and time-consuming tasks. Hence, optimizing image-processing tasks based on a few gold samples is a highly desirable task. This paper demonstrates how combination of genetic algorithms and morphological operations can be used to generate an image processing procedure for object extraction. For this purpose, the approach receives the original image and a user-prepared image as gold sample reflecting the useršs expectations. After carrying out the optimization phase, the optimal procedure is generated and ready to be applied on new images. The approach architecture and the employed methodologies will be explained in detail. Experiments have been performed on grey-level images and results will be provided. |
| |
|
| | |
| | |